Jump to content

Broadwell with Intel HD5500 and Intel Core i5 5200U Framebuffer's For Devices> Properties --Clover


4 posts in this topic

Recommended Posts

This may help those having to slap the table, throw your wife out the window, curse at the cop, and what not that may probably occur

after hours, if not weeks of hours to wondering, why the @@#$ doesn't this work as Recommended!?!?!?!

 

After much research trial and errors' got results, like a lab rat does.

 

I came up a working solution to an issue which I had with the Default Recommended settings that did not produce the desired results.

Encountered a situation where the Recommended settings did not work for the HDMI portion at all.

From having no sound and no video output. Even for non-HDMI sourcing due to, not sure. But a bug for sure.

 

Working with:

HP Probook 450 G2 model year 2015

Intel Core i5 5200-U, HD5500 IGPU and 9 Series North and South Bridges.

oob ig-platform-id is 00001616, or 16168086 or 80861616 whichever you prefer.

 

Using the 'Recommended' by OpenCore Guide for Broadwell Laptops suggests to use 06002616 and Whatevergreen FAQ page for Intel IGPU's.

Also, having to resource other guides to put together the full chapter. --> pages missing and gaps create issues.

Both ig-platform-id's work and each one produces different designations as seen in Hackintool.

Thus the need to write into Devices> Properties the actual frame-buffers to get the desired results, ie., as con1 and con2 with alldata.

MacOS Monterey 12.3.1.

Clover Bootloader 5146

 

Beginning the process:

 

According to the WEG FAQ page for configuring the Intel integrated graphics or IGPU, eGPU, eDP,  to work with Whatevergreen.kext is written on the linked site below.

https://github.com/acidanthera/WhateverGreen/blob/master/Manual/FAQ.IntelHD.en.md

 

According to what is written and it may be very gappy for some cases and not enough information in others, but a good starting place.

 Find the appropriate, PciRoot designator for your graphics hardware device and write the following:

 

Devices>Properties

PciRoot(0x0)/Pci(0x2,0x0)

AAPL,ig-platform-id: Data : 06002616

found here:

https://dortania.github.io/OpenCore-Install-Guide/config-laptop.plist/broadwell.html#deviceproperties

 

With a written on that page, recommended Framebuffer configs from OpenCore Guide for stolenmem or cursorbytes:

 

Too small:

framebuffer-patch-enable Data <01000000>

framebuffer-stolenmem Data <00003001>  19MB stolen (framebuffer-stolenmem)

framebuffer-fbmem Data <00009000>  9MB FB mem (framebuffer-fbmem)

 

Framefuffer findings from Kext in Apple /S/L/E  ( Herve posted here in forum )

 

framebuffer-patch-enable Data <01000000>

framebuffer-stolenmem Data <00002002>  -> 34MB stolen mem (framebuffer-stolenmem)

framebuffer-fbmem Data <00005001>   -> 21MB FB mem (framebuffer-fbmem)

 

And that's it nothing else but,

 

According to the WEG FAQ Intel.en.md page:

For connector 00040000 WEG would inject 0008000 to get HDMI Audio.

 

WEG would take care of the patching for supported hardware. This Broadwell is definitely supported.

Look in /S/L/E to find if your graphics is supported with kexts.

Such as AppleIntelBDWGraphicsFramebuffer.kext  

Screen shot 1

 

here's a noted portion from the said page we use; look and read the last portion of this paragraph below which begs for a layman's term on the WEG FAQ.IntelHD.en.md

Such as: Your HDMI needs the correct !^$&*#$ frame-buffer written in the config man! with so you can hear what is being output.

 

Digital Audio (HDMI / DVI / DP)

 

To enable digital audio it is necessary to set the hda-gfx properties and patches the connectors.
To enable audio in general and HDMI in particular use WEG along with AppleALC.kext. AppleALC automatically injects missing hda-gfxproperties.
On 10.10.5 and above, WEG automatically changes the connector-type of DP (00040000) to HDMI (00080000), only if not used Custom patching. Physical connection may be of any type (HDMI, DVI, DP), but for the digital audio connector-type must explicitly be HDMI. --courtesy of the OC team! great work!!

 

What this means is, for HDMI Audio to output to your TV or sound system, you need to write in the proper config designations or no $(*Q*(##$^ audio.....

 

What had happened with this Laptop:

 

After reading this last portion of the Paragraph above,

Have to set an HDMI manually for the < ig-platform-id of 06002616 > due to con1 and con2 being  in reverse.  Apple and Windows do not use the same

ports for the same design.  Apple says, HDMI audio is on con1, Windows says, HDMI audio is on con2.

but it's actually Intel who makes it that way

for PC Windows and Apple MacOS. So Apple reads it one way and Windows reads it is another. Confused yet? 

 

This is what WEG defaults to with no Framebuffers configured in config.plist in Devices> Properties: and can be confirmed with Hackintool:

 

01050B00 00040000 07050000  -> port #2/FB@1: index 01, busid 05, pipe 0B00, type 00040000=DP  ((should be DigitalDVI))     (framebuffer-con1-type), flags 07050000

02040B00 00040000 07050000  -> port #3/FB@2: index 02, busid 04, pipe 0B00, type 00040000=DP  ((should be HDMI))     (framebuffer-con2-type), flags 07050000

 

With the above configuration no HDMI Audio coming out from con1 or con2. 

 

What was happeing:

With just the stolenmem framebuffer patching in Devices> Properties, result was HDMI video, but no HDMI audio.

With just the AAPL,ig-platform-id data <06002616> as recommended, no HDMI audio. and funky screen position when HDMI sourced.

 

Once you enter into the configs, in it,  it is considered 'Customized.' So adding a stolenmem patch starts the process to Customized.

 

This may or may not  be true for other laptops but for this one, this is the case.

 

So we go back to basics we start with the board or IGPU's of this boards' ig-platform-id to get an output in English such as

This con is DigitalDVI, This con is HDMI, This con is Main Display, your face and eyes are staring at.

 

Working with the Stock ig-platform-id.

 

Changed the Device Properties from:

 

AAPL,ig-platform-id     TypeComment

06002616   Laptop   Recommended value for Broadwell laptops

 

To: 00001616

 

the output below  is per Hackintool, this is what Hackintool sees:  [ No frame-buffers are written into Devices> Properties, PciRoot ]

 

bus 0 -> con0  00000800 02000000 30020000  LVDS Display

bus 5 -> con1 01050900 04000000 04000000  DigitalDVI
bus 4 -> con2 02040900 00080000 82000000  HDMI ( but no sound whatsover.) 

 

also added con0:

 

00000800 02000000 30020000  LVDS — main Display

 

Changed the con1 and con2 around like this:  changed where HDMI Audio comes out from, the output of sound.

 

01050900 00080000 82000000  HDMI Audio sound working and video

02040900 04000000 04000000  DigitalDVI <> don't have a cable to test

 

Now this is quite backwards and probably other people may have similar issue.

This laptop being IBM based and Apple being Apple Based not the same animal. Flipped or Reversed in order for the con's, such as con0, con1, con2.

Configurations do not match in some cases as recommended must work to find what works.

 

Went back with Recommended by OpenCore Guide:

 

So this is the current and HDMI video and audio working. Stolenmen working with higher numbers.

 

AAPL,ig-platform-id 06002616

#device-id 16260006 < omitted, if not omitted have no desktop mode >

framebuffer-con0-enable 01000000 < not needed as WEG handles this and confirmed in Hackintool can be cosmetic >

framebuffer-con1-enable 01000000

framebuffer-con2-enable 01000000

framebuffer-con0-alldata 00000800 02000000 30020000 < not needed as WEG handles this and confirmed in Hackintool >

framebuffer-con1-alldata 01050B00 00080000 82000000 < HDMI audio and video >

framebuffer-con2-alldata 02040B00 00080000 82000000 < should be DigitalDVI and probably be switched to 02040B00 00040000 07050000 if no DigitalDVI >

framebuffer-patch-enable 01000000

framebuffer-stolenmem 00002002 < 34MB

framebuffer-fbmem 00005001 < 21MB

 

Conclusion have to do the work.

 

 

 

 

 

 

 

 

Screen Shot 2022-04-01 at 6.56.08 PM.png

  • Like 1
10 hours ago, Hervé said:

OMG! 😱 😳 😱

 

There'd be so much to say... and it certainly ain't the wife I'd throw out of the window, it's the thread.

 

Let's try and put right the numerous incorrect statements and falsehoods you made (though some are very funny):

  • "This [HP ProBook] laptop being IBM based and Apple being Apple Based not the same animal.😂
  • " From having no sound and no video output. Even for non-HDMI sourcing due to, not sure. But a bug for sure." Aye, a big 2-legged one, I reckon. 😂
  • it must still elude you that it's an Intel standard that iGPU is @2 on the PCIe bus on all Intel platforms (and dGPU is @1 on an Intel mobile platform). Hence why all iGPU properties are always injected against IO location PciRoot(0x0)/Pci(0x2,0x0) across the board. There are no exceptions to this.
  • you appear to confuse many many things from framebuffer layouts to iGPU ids, through vendor/device ids, endianness or framebuffer connectors.
  • there is no such things as ig-platform-id 16168086 or 80861616; these would be entirely invalid. 8086 is Intel's vendor id (in hex); see below for 1616.
  • there is no such thing as iGPU device-id 0x06002616 (or 0x16260006 to that effect). If you inject this, of course you'll have no graphics acceleration at all ! 😲
  • i5-5200U is fitted with HD5500 iGPU which carries id 0x1616; it is natively supported by OS X/macOS (cf. BDW drivers) and therefore requires no injection.
  • OS X/macOS graphics framebuffers are drivers written for Mac computers, not PCs that owners wish to run as Hackintosh; so of course they no not match the Wintel specs.
  • connectors are video output ports and there are no such things as "flipped or reverse" connectors. "Swapping connectors" as you PM'ed me is a non-existing/invalid concept. You patch/redefine/reconfigure them. Sometimes, you even add or remove connectors.
  • a Broadwell laptop with a mainstream 5th gen. Intel Core i3/i5/i7 "-U" CPU and Intel HD5500 graphics is meant to call on BDW framebuffer 0x16260006 and SMBIOS MBP12,1. Forget about desktop layout 0x16160000, it's not suitable for your laptop (and bears no relation to mobile iGPU id 0x1616).
  • OpenCore team did not reveal to the world that patching the (appropriate) connector to HDMI type 00080000 was required to gain HDMI audio; that's been known for the best part of 10 years and pre-dates OpenCore by many such years.
  • It's commonly known that HDMI output port usually registers against connector con1 on Hackintosh laptops. To gain HDMI audio, all you have to do is set that connector to HDMI type by injecting corresponding value 00080000. There is no need to modify the connector's flags as you did (the result of your erroneous belief that connectors con1 and con2 had to be "swapped").
  • of course, there's no need to modify connector con2 to gain HDMI audio if HDMI is attached to con1. con2 may attach to a DP/mDP or DVI output (even through a dock) and may (or may not) require to be patched accordingly.
  • as for framebuffer patching re: DVMT, I believe the detailed explanations I posted here should be clear enough. But I sure don't know what you meant about me posting about "framebuffer findings from kext in Apple S/L/E".
  • I've stated it dozens of times, injecting properties that are identical to a framebuffer's native properties is utterly useless though it does no harm of course. Consider it the same as adding a "Ford" sticker on all doors of your Ford automobile to indicate it's a Ford when it already has a Ford logo on its tail end and on its bonnet/front grill.
  • your injections setting stolenmem to 34MB and fbmem to 21MB fall into that category of useless injections for these are the values defined in BDW framebuffer layout 0x16260006:
ID: 16260006, STOLEN: 34 MB, FBMEM: 21 MB, VRAM: 1536 MB, Flags: 0x00000B0B
TOTAL STOLEN: 56 MB, TOTAL CURSOR: 1 MB (1572864 bytes), MAX STOLEN: 124 MB, MAX OVERALL: 125 MB (131608576 bytes)
Camellia: CamelliaDisabled (0), Freq: 2777 Hz, FreqMax: 2777 Hz
Mobile: 1, PipeCount: 3, PortCount: 3, FBMemoryCount: 3
[0] busId: 0x00, pipe: 8, type: 0x00000002, flags: 0x00000230 - ConnectorLVDS
[1] busId: 0x05, pipe: 11, type: 0x00000400, flags: 0x00000507 - ConnectorDP
[2] busId: 0x04, pipe: 11, type: 0x00000400, flags: 0x00000507 - ConnectorDP
00000800 02000000 30020000
01050B00 00040000 07050000
02040B00 00040000 07050000
  • If those default values work for you, I guess you must have missed that your laptop probably has DVMT pre-allocated memory set to 64MB or higher; which means you don't even need to patch fbmem and stolenmem in the 1st place, let alone inject the same values as defined by the framebuffer! -_- It certainly would have resulted in KP had your laptop been set with 32MB DVMT pre-alloc mem.

Until you actually better and properly understand the way things work in macOS, I would advise you to be more cautious about what you post (I refer to the numerous posts you made recently).

 

All you'd have to inject in your bootloader configuration should be:

ig-platform-id           06002616    DATA
framebuffer-con1-enable  1           NUMBER
framebuffer-con1-type    00080000    DATA

// Optional depending on Broadwell laptops's DVMT default settings but usually/often required
framebuffer-patch-enable 1           NUMBER
framebuffer-fbmem        00009000    DATA (or whatever as long as fb+stolen < DVMT pre-alloc mem)
framebuffer-stolenmem    00003001    DATA (or whatever as long as fb+stolen < DVMT pre-alloc mem)

 

Ok, you discover this only in 2022 but rest assured that there are literally hundreds if not thousands of threads/posts providing this kind of information since support for Broadwell was introduced in Yosemite back in early 2015.

Thanks Herve

 

this is what I needed the actual config.

 

Like you posted exactly what is necessary for supported for this unit in a few strokes.

 

It's not that one doesn't fully understand it's what is necessary to use for the instance.

There are countless information I'm certain that is written but many are not fresh and maintained

as if it was 10 years ago due to the newer information and data collected.  Archived. common problem.

 

Like a few years ago Clover was 4000 before that 3000. Today 5000.

MacOS was Mojave 10.14.  To keep up with the changes there are tons of changes made

and which one is correct?  So to have someone like yourself who maintains constancy through

the changes to know what has changed vs what is constant to configure for that instance.

 

But sifting through the information such as a person who doesn't spend the days, collecting every bit of information,

but rather just wants to get there unit up and running, there is too much to go through ie., the changes.

Unfortunately as time passes, these get archived or deleted. What was fresh then and there.

What was evidently there yesterday is not there today at your finger tips as in right there.

and is it thousands of pages to sift through.  Where as for someone like yourself and others like yourself

who maintain these, for these it is relatively fresh data as compared to someone like myself who has

to collect the data to sift through to understand what is good vs what is no good.


Sifting through old data to find the correct data and to know if this data is viable for that particular instance,

to know for certain.  Then you have to find if that data is correct that person has written, so the search

is continuous. Who is correct and who is half correct? who knows the exact? Obviously someone who

 spends their time in it.

 

As I pointed out up top in my post which set of configs is to be used for this instance?

This person has all the information written but this posting is general and few examples.

So finding the correct example that fits that instance.

 

I'm glad you had some laughs, because also there are plenty of laughable things to consider from

those outside peering in.  Human habits.

 

And thank you for setting the correct information. 

 

Perhaps the words are necessary to provoke one to getting answers.

 

LOL!! haha

 

 

Edited by makk

Personally I like OpenCore because they have a guide you can use to configure which is nice to have because for someone like myself who 

wants to get Hachintosh up and running in they pretty much make one feel 'confident' in doing so.  Catch word 'Confident'

 

There are no egos on that guide.. They do their best to maintain professional attitude to present the facts needed.

 

Although it begs to have it more detailed pertaining to each instance rather than a broad general coverage.

 

Makers makes changes in the hardware such as Asus, Acer,  they do things each their own way even though they have the same hardware in the units.

Benchmarks tell the tale.  Exactness does not occur. Asus and Acer do not produce exact copies of each.

 

Intel HD 5500 in an Asus is different from an Intel HD 5500 in an HP. 

The chip is the same, integrated in the CPU but what they do it is different in the flash and other things. A standards must be kept

but who keeps standards perfectly?  No one.

 

So we are in thought looking for the exactness for that instance at that time and not 10 years ago

And not everyone is abreast of Broadwell, Kaby Lake, Ice Lake, Comet Lake, Sandybridge, because it is quite not

a subject that is going to keep your life together It is an hobby or some lower interest. Not high on the totem pole.

 

So I will find a way to run Windows from Inside OpenCore so I can maintain some sanity compared to the long set of configs

in Clover which there is no Guide like OpenCore has.

 

Anyway thanks for your time. Have a good weekend 

 

 

 

You just proved my point about information.

 

This is an HP Probook 450 G2 laptop with an ig-platform-id or Intel '8086' which are all based on IBM principles. IBM being the one to introduce Personal Computers.  HP having history with Unix. 

 

ig-platform-id when used in retrieving data for to run a Macky hacky purpose, hackintosh, hachintosh, jakintosh, whateverintosh.

 

for PC's it's another lingo. PCID, DEVID, PID, all the little acros that geeks use.

 

ig for this one is 16168086  Laptop. There's no delimiter here, it's a freaking Laptop with that set.

So for Apple to understand however you interpret dialogs and words, it has to be' converted, or swapped or reversed or, whatever

you like to name it, so this Laptop will boot and run.

 

Conversion tables, are inconsistent, lacking professional albeit, good Samaritan, along with a great perfect set of documentation that is 'Constant' through the years. No archiving to the back row dungeons.

 

case in point, there are plenty of older Architecture that people still run or install MacOS on.  They need these  constants.

They are all kinds, and categories of users.  Thoughts, intents, interests, so forth, not just those who do this all day long, eat, sleep and breathe.

 

So if the tables, indexes are constant, then most will be on the same page, sentence, paragraph.

 

Thus it is  and thus it will not ever change because no one wants the headache, so they keep their little corners and ride that.

Which is small minded not broad and covering.

 

Thank you for your input.  It would be better if there were actual constant tables or indexes with the pertinent information

so a baby can install and modify their IBM based computers.  Interesting point we all most of us went through education

and needed teachers to instruct and help start a database to go along in life.  To teach and instruct.

Data is a broad and massive industry beyond one's scope because it encompasses infinite subjects and encapsulates all the conditions

in life to it.  No one person can know the infinite subjects, unless they were Infinite themselves.

 

 

 

×
×
  • Create New...